sensitivity question


A few questions regarding sensitivity.
a speaker with 90db sensitivity means roughly that it will deliver 90db at one meter distance at 1W right?
But if I’m not wrong most speakers are emitting more sound towards the drivers direction. In some way they resemble a light bulb+projector which emits light toward one specific direction. Obviously with speakers the "projector" effect is not as pronounced as with a lighting projector however the sound is mainly emitted toward one specific direction. With light bulbs which have a projector, in order to describe correctly the emitted luminous flux, you have to add the light cone degrees. In fact a spot light which opens one degree and consumes one watt delivers at one meter more lux as a light bulb of 100W at one meter which opens 180 degrees.
Wen measuring the sensitivity of a speaker do they measure only the db at one meter from the front of the speakers or do they make 8 or more measurements at 1 meter around the speaker and calculate then the medium sensitivity? If not is the actual sensitivity measurement not strongly lacking in accuracy? And in a audiophile world which has measurements in such a high regard is that lack in accuracy not wired? Thanks for any info!
128x128daros71
A few questions regarding sensitivity.
a speaker with 90db sensitivity means roughly that it will deliver 90db at one meter distance at 1W right?


You are almost correct. :)  Sensitivity is output vs. 2.83 V, regardless of Wattage.   If your speaker is exactly 8 Ohms then 2.83V would correspond to 1 W.

You are correct that this measurement is from the front of the speaker.  This measure is somewhat subjective as you must eyeball around 1 kHz.

The dispersion of a speaker is a complicated thing, and even within the same loudspeaker it varies greatly from driver to driver and depending on crossover design.  If you are truly curious I suggest you look at the Stereophile speaker measurements for the off-axis measurements.

I think you are asking for sensitivity to do much more than it is intended.  It's basic function is to explain how loud it will be vs. a reference input.  What you seem to be trying to do is use this 1 number to explain the 3D soundfield produced by a speaker.  No, it is not weird.  Sensitivity and efficiency each do what they are intended to do.

If you want to understand the off axis response of a speaker you'll need different measurments, such as what Stereophile provides, and if you want to now how it will perform in your room, well that's something else entirely.
Thanks for the exlpaination! This means that lighting fixtures are measured more accurately as speakers? Funny. Now i start to understand from where all this subjectivity cult in the audiophile world comes from.
You can make polar plots of a loudspeaker's output but output generally is measured on axis. To get total output is a complex math and measurement problem but since that is not the industry standard, it isn't used.  
Thanks for the exlpaination! This means that lighting fixtures are measured more accurately as speakers? Funny. Now i start to understand from where all this subjectivity cult in the audiophile world comes from.


That's' a ridiculous leap.
This same guy that start butcher thread with speaker like banjo. He is try to make reaction because he don’t like audio. He should try Maggie speaker and see what real man speaker can do not some cheap junk.
It's getting a little better. Look for measurements using the klippel NFS. I believe Stereophile is going to start using one. 
I will say this regarding speaker specs- I recently auditioned two speakers in my home- one manufacturer (Wharfedale Linton) stated the sensitivity at 90 db, the other (SVS Prime Pinnacle) 88 db. The SVS played louder for a given volume level, and the meters on my amp confirmed this. 
I have a relatively low powered Luxman L-590 AX MK2 amp (30 wpc). The supposedly lower sensitivity SVS speaker did not move the meters for a given volume nearly as much as the supposedly higher sensitivity Wharfedales. In fact, the Wharfedale came close to redlining the meters at high volume, the SVS not even close.
This admittedly non-scientific trial tells me not to put much stock in sensitivity ratings. 
It may somehow dovetail with your lighting analogy- I’m not sure though. Both speakers are dynamic, front firing boxes.
I bet you put you little Peter behind the lamp light so it look bigger!!!
Good answer Erik.
10 ft. pole too short for me.
You make sense.
The OP don't.
That’s’ a ridiculous leap.
I think the sensitivity of a speaker, as measured and stated by speaker manufacturers, is only appropriate for headphones. For speakers it is 100% wrong. It simply doesn’t describe the amount of db reverberating in your space. If you measure something and you measure it incorrectly, that measurement won’t help you describe the world objectively... and then, since we all have different experiences despite the same measurement, you’ll say it’s all subjective in the end.

P.S. i don't say that a correct sensitivity measurement will reveal the overall qualities of the speaker. I want only to point out that in hifi even the simplest measurementes, like the sensitivity of a speaker, are gathered in a very discutible (or subjective) way. But is a subjective measurement useful? 

Subjective if it is, subjective. It's not. The criteria for accurate measurements only need be noted as a 4 6 or 8 ohm voice coil.

WHERE the mic is placed and the fact that it is calibrated. That is the data we look for. The way to take measurement is to note HOW you did it. Should note temps, humidity and any other features. LIKE the ROOM.
The best place is at night in the open back yard...

The space the driver is in has nothing to do with it's POTENTUAL to perform in ANOTHER given space.. I think you're mixing up what a driver can do with what a drive IS doing...

Erik explained off axis response is just that. The better question is does it effect YOUR sweet spot because of it or the lack of it.. Off axis response that is...  Wouldn't YOU have to be OFF axis to hear the difference. No ones moving in a listening room.. If they were it would be a BALL room. AY?
The reasoning here is nuts.

To make a car analogy, they measure the MPG of the car and you are upset it doesn’t tell you if the car has a hatch back. You are fixated on a measure that does exactly what it should do (and does well) because it doesn’t do what you want it to do and no one, I mean no one, intended it to do.
Are you sure? This guy for example( https://geoffthegreygeek.com/understanding-speaker-sensitivity/ ) is stating that the sensitivity measurement should describe how loud a speaker will sound given a certain input. I only say that this is wrong, that the sensitivity, as it is measured now, do not describes how loud the speaker will sound until you do not include the off axis sensitivity. Moreover a speaker which has a 20 degree sweetspot (i’m not sure if the sweetspot is the correct indicator) which measures 90db sensitivity will sound less loud as a speaker which measures 88db and has 60 degree sweetspot because the second one simply moves much more air, which soon or later, after reverberating in the space, will reach you ears. Only in an anechoic chamber you will hear this 90db at 20 degree sweetspot speaker sounding louder as the 88db 60 degree sweetspot speaker. Buth woh is listening music in such circumstances? No one. To me the actual sensitivity standard is very blurry. But if you are happy with it...
My favorite amp is 1 watt. Should I look at the thickness of the wood to find the best speaker. How about the shape of the feet? I wish I knew a common way to measure them? 
I will attempt to break it down to understandable terms.
Through the years, the standard for rating speaker sensitivity has always been how loud a speaker plays with 1 watt of input power with a microphone measuring 1 meter away (on axis)
This is a VERY inaccurate way of measuring. Why?
Impedance is not flat, that is why we call them impedance curves. I have seen speakers rated at 8 ohms have 3 ohm dips...
Remember Solid State amplifiers normally put out more power at lower impedances. (tubes are more consistent) So to get consistent power output to a speakers varying impedance can be tough.
So, yes 2.83 volts is 1 watt to 8 ohms, but 2 volts is 1 watt to 4 ohms and 4 volts is 1 watt to 16 ohms.
So as you can see, applying exactly 1 watt to any speaker is nearly impossible as impedance strays.
This is why I believe their should be a standard for all speakers to be measured at 2.83 volts regardless of impedance.  With this,  we will know how the sensitivity any speaker compares with a reasonable degree of accuracy.  The only thing that we would have to make sure of in an amplifier purchase would be: Can my amplifier drive this speakers impedance curve or for tubed amps (capacitive phase angles).
In my speaker builds, I believe in impedance compensation circuitry. This has always produced a better sounding speaker for me and a much more stable load, which amplifiers seem to appreciate also.